To: Emelie G.
From: Geoffrey Klempner
Subject: Fairy tales and the coherence theory of knowledge
Date: 10th November 2010 12:27

Dear Emelie,

Thank you for your email of 1 November, with your essay for the University of London BA Epistemology module, in response to the question, ''The fact that there are coherent fairy tales shows that one might have a coherent system of beliefs without those beliefs amounting to knowledge.' Discuss.'

This is a good essay which is well up to the standard of work produced by my other University of London students.

A great deal of work has been done on this topic by contemporary epistemologists, and an examiner would not expect you to be able to summarize all the various positions right up to the current state of debate. For example, you don't discuss the contrast between 'internalist' and 'externalist' accounts of justification in terms of coherence. That would possibly be a relevant consideration, but, as I said, you are not expected to summarize the entire debate given the limited time available for answering the question.

What are you expected to do? Answer the question. That entails saying enough to show the examiner that you are fully aware of exactly what the question is.

Here, arguably, you have been a little imprecise, in seeming to conflate a coherentist theory of knowledge with the coherence theory of truth. Suppose the question had been, ''The fact that there are coherent fairy tales shows that one might have a coherent system of beliefs, many or most of which are in fact false.' Discuss.' That is an objection targeted against a coherence theory of truth. You cite Brand Blanshard who is perhaps one of the best known defenders of a coherence theory of truth. However, one might question whether a response to the objection against a coherence theory of truth, along the lines adumbrated by Blanshard, is needed in the case of a coherence theory of knowledge. It is perfectly possible to hold a coherence theory of knowledge while at the same time rejecting a coherence theory of truth.

Another point at which we need to make an important distinction concerns the aim of a 'theory of knowledge', in the sense of a definition of 'S knows that p' in terms of necessary and sufficient conditions. There is no necessary entailment between a definition of knowledge in this sense and an adequate response to the full-blooded sceptic. Suppose we had an adequate answer to the objection to coherentism about knowledge posed in this question. One is not thereby armed against any or all objections that might be raised by a full-blooded sceptic. The response to scepticism might require a different strategy, perhaps a strategy on which holders of competing definitions of knowledge in terms of necessary and sufficient conditions could agree.

Why be a coherentist about knowledge? You give the most compelling consideration: that coherentism is a reaction to the implausibility of foundationalism. There just aren't enough foundational beliefs around on which to base the edifice of knowledge. (You could have added that this is much more evident given that we have rejected traditional foundationalist theories which identify foundational propositions as those concerning, e.g. sense data.)

If that's true, however, how can we fail to be coherentists of one kind or another? We have to accept that tracing back the regress of explanation/ justification will eventually take us in larger or smaller circles. Why is that so bad?

I think here that, without conceding the case to the contemporary foundationalist, more can be said about the structure of knowledge. Not all beliefs are the same. In particular, special considerations apply to beliefs based directly on experience -- notwithstanding your example of Jane who 'insists that she sees flying green horses in the sky.' Do we just take her word for it? Are we stuck at the point of insisting that *we* don't see them, even if she does? But, surely, there is much more to say about *how* we perceive. Suppose we ask Jane to take a photograph of the green horses that she sees. Then we look at the photograph together. How many horses are there? Exactly what shade are they?

Suppose one accepts that, in principle, a sufficiently resourceful Jane could weave a wonderful story justifying every claim that she makes, despite all the seeming evidence to the contrary. (I once had an evening class student who believed that the earth was flat.) Isn't this really a worry about scepticism, rather than a worry about the concept of epistemological justification? Accept that the flat earther, or the green horse believer, cannot be persuaded by any argument, no matter how ingenious. I'm not going to have any sleepless nights worrying whether maybe the earth is flat or that there really are green horses that for some reason I can't see. Accepting the existence of flat earthers or Janes is consistent with holding that the best analysis of justification that we have necessarily involves appeal to coherence.

I have been talking (as in fact you have) as if foundationalism and coherentism are the only alternatives. Your brief mention of Gettier, however, suggests further possibilities such as a causal theory. Why can't we simply take all the best bits from all the theories and say that there are relatively foundational propositions (though not enough of those); that coherence and causation are also involved? In that case, the objection in the question can be accepted as a rejection of an overly simplistic account of coherence, but one which is easily met if we are prepared to complicate our theory a little.

All the best,

Geoffrey